New clinical tool helps predict short-term risk of diabetes complications using routine health data

Researchers at the University of Maryland School of Medicine have developed a new risk calculator that can predict an individual's short-term risk of developing various diabetes-related complications using routine health data. This tool, known as the Diabetes Complications Risk Calculator (DCRC), can estimate nine different acute and chronic complications simultaneously and update these estimates over time as new clinical information becomes available.

The study, led by Dr. Rozalina G. McCoy, analyzed health data from over 400,000 adults newly diagnosed with diabetes across the United States. Unlike existing tools that focus on a single complication or predict risks over extended periods, the DCRC can assess the likelihood of multiple complications within a month and adjust these estimates based on evolving health status.

The calculator incorporates commonly available information like age, existing health conditions, medications, and laboratory tests. In testing, it showed good to strong accuracy in predicting complications both in the original nationwide dataset and in an independent group of patients treated at Mayo Clinic.

Within one year of diagnosis, about one-third of patients experienced at least one complication, rising to over 40% after two years. Factors linked to higher risk included older age, existing cardiovascular disease, and certain medications. The model also identified trends where risk could fluctuate as health conditions and treatments changed over time.

While the DCRC demonstrates the potential of machine learning in routine clinical settings, it's important to note that the tool should be used in conjunction with clinical judgment. Further testing is needed to assess its performance in everyday clinical practice and to determine its impact on shared decision-making and long-term health outcomes.